Bhaskar Anepu, Co-Founder and CTO of Smirta Innovations
SMIRTA has 2 companies under its purview. How did that come about and what is SMIRTA’s role with both of them?
I’ll start off by saying that I have over 15 years of experience in software development that eventually evolved into innovation and systems project management. I last worked for the mobile division of Intel, but I always wanted to get into entrepreneurship. My wife is an ER physician, so whenever she comes home, there’s always a pain point that we discuss. I realized that there’s something we can do there that can help the healthcare industry. That’s how the SmartER-Care product came about, and it was the first product that SMIRTA started to work on.
KidneyAI was the second project. I knew Dr. Kolachalama for several years, and when I decided to leave Intel to work on my company, he pitched an idea that entailed developing machine learning models to deal with chronic kidney disease patients. The premise was that machine learning would make the process more efficient, not just clinically, but also for the pharmaceutical industry. The idea was very promising, so we embraced both ideas under SMIRTA. Dr. Kolachalama’s lab came up with the models and technology, and we’re using SMIRTA to commercialize it into something that’s usable for the industry.
Is there a reason for combining these two entities under one umbrella as opposed to keeping them separate?
They are separate entities though the umbrella is the same. However, at the same time, there is a slight overlap because smartER-Care will also be using data over time to apply machine learning to do some predictive analytics for urgent care centers and emergency rooms (ERs). At this point in time, there is a little overlap, but they’re both healthcare ventures.
Who do you anticipate being the end user of the kidney AI technology and how are they going to use it?
Kidney AI can serve hospitals, nephrology practices, diagnostic labs and transplant centers in a clinical setting. First, nephrologists can use the software to quickly identify fundamental markers in kidney biopsies. In hospitals, the nephrologist can use these markers to make quick decisions before a complete diagnosis can be made from analyzing the biopsy. The second important indication is for kidney transplant centers since the decision on whether or not a kidney is appropriate for transplant has to be made very quickly. Nephropathologists usually are very rare, so access to a general pathologist for kidney transplant biopsies is the current practice. A solution like KidneyAI would be more targeted to the problem by allowing the decision on whether or not the kidney is usable for transplant to be made quantitatively and objectively, saving a lot of wasted kidneys.
In addition to these clinical settings, KidneyAI can also be used for drug development. Contract Research Organizations (CROs) and pharmaceutical companies use biopsies to study the impact of drugs on animals and humans. Here, the same principle can be applied. Ultimately, we need to analyze the biopsy using our software, but the impact is different because it’s used specifically for drug development. In preclinical trials, this technology can be used on animal biopsies to see if the drug is performing as intended. These are not just limited to kidney-related drugs since many other drugs can have side effects that negatively impact renal function. In pharmaceutical and CRO settings, our product will help speed up the drug development process and reduce cost to market by making it more efficient in patient selection for clinical trials, for example.
Where did you get the data to train these models?
Dr. Kolachalama who is an Assistant Professor at Boston University School of Medicine (BUSM) used the data from Boston Medical Center (BMC). Our work is published based on years of research at academic labs in large US medical centers.
Where does the data come from for drug companies that are running studies on animal models?
The model is going to be similar to the human model but must be trained on animal specific kidney biopsy data.
Switching over to your other company, SmartER-Care, what is your business model? Who is paying for the service you are providing?
SmartER-Care targets three different entities, one of them is patients, who are the most important for us, for whom it is free. The product also significantly helps payers, who are health insurance companies, and hospitals that have emergency rooms and urgent care centers. For hospitals and urgent care centers, we have a tool called SmartER-Care PRO, which is another part of this platform that is different from the patient facing SmartER-Care mobile app. There is a touchscreen kiosk and a dashboard that we deploy at urgent care centers and ERs that staff uses to monitor and manage patient flow. The app and the kiosk are parts of a framework that has multiple moving parts. These services provide revenue from hospitals, where it helps them with patient engagement and other factors. We also have insurance companies and employers (since directly or indirectly they are the payers for health care). For them, it helps since it is not a generic ER or urgent care finder. Instead it shows results based on a patent pending algorithm that uses patient information and provider details. While the app is transparent to the user, behind the scenes there is an algorithm that was designed by ER physicians to determine with high probability if the problem the patient is experiencing requires an ER or can be handled at an urgent care center, which is much less expensive. In some cases, the insurance companies are refusing to reimburse ER patients since they determine that the problem could have been dealt with at lower cost at an urgent care center. This is unfair since generally patients are not educated enough to make that medical decision. Hospitals are facing a problem since someone with a minor injury gets upset after waiting for 2 hours behind more life-threatening patients. Our product helps all these three entities, and while the patient is the main person we want to help, the insurance companies and hospitals also benefit significantly in terms of reducing their overhead costs and giving better patient experience. So that is where the revenue comes into the picture. All stakeholders will be interested in using this.
Did you try to integrate this with existing EHRs such as EPIC or is this a stand-alone service?
We provide a stand-alone solution, but we are also working with an urgent care center where we are integrating it with their EHR as we speak.
How are you marketing this product to your end user, patients?
That is a very good question because this is not a video-sharing app, or social media app or a game so it is not easy to market. One of the approaches we are thinking of is signing up health systems, employers and insurance companies. We can reach out to their employees, patients and clients. At that point it may not be standalone, it may be integrated into each of their own apps since pretty much everybody has an app today. So, this can be an add on to those or be channeled to their end users as a standalone app. The other approach is through typical means such as advertisements on social media platforms, professional platforms like Linkedin and having trifolds and flyers in physician group offices, urgent care centers and ERs. We are taking a multipronged approach to target the end user, but it is definitely one of the more challenging aspects considering this is not an app someone is going to use every single day. Although ultimately the roadmap is going to lead to a point where we integrate all of someone's health care scheduling and management into one place.
How do you see this technology evolving and what do you see this app looking like once it comes to fruition?
Ultimately it has to go beyond just urgent care centers and ERs, because the goal of this is the right care, at the right time, at the right place. So that is not restricted to urgent care centers and ERs. Today, improving ER care is our primary target due to the high costs and multiple pain points. But we plan to integrate other physician offices as well so that you do not have to wait 30-40 minutes to see the doctor. That does not make a lot of sense since the whole point of an appointment is that you will be seen within a reasonable amount of time around that time. That is why we plan to expand this to other care providers such as Primary Care Physicians (PCP). Additionally, any visits to an urgent care center or ER would automatically be visible to their PCP. This would also be true of specialized doctor visits, surgeries, and telemedicine appointments. Patients and insurance companies are going to prefer and push towards telemedicine over costly urgent care center or ER visits. All of this information is going to be accessible from the app and will act as a central repository for the physician and the patient. Today, even though the whole concept of EHRs and EMRs is to make this information seamlessly available to different health systems, that is far from the fact. So, if we can create an ecosystem that I am talking about, then you are connecting multiple health systems via the patient.
How is HIPAA compliance and security being taken into consideration with all of the data that is being shared? Also, who will own the data once it is being collected?
HIPAA compliance is needed even today when patients send their symptoms to urgent care centers, so we are already following HIPAA compliance. But, when we go to the next level as I just described, our first attempt will involve giving access to patients via our app to the data across all their affiliated providers. So, we don't want to “own” the data, but we want to create connectivity between the patient and these (systems) all in the same, unified ecosystem. So, for the end user, when you go into your profile and you want to look at all your prescriptions and PCP visits from the past 6 months, and say, you had a surgery in a totally different health system, all of these will be “accessible” from your profile. However, we will not host the information and we wouldn't want to own the information. We are simply providing the connectivity between the place you had the surgery and you. So, you are actually accessing the information from them, but we make it seamless for the patient. Similarly, all your PCP visits will be connected to you as well. So, we would not want to own any of the information or put it on our servers, instead, we would simply want to provide you a means to access all of your information from one place.
That being said, as we approach this, we’ll have to see how it goes because we want to make it as practical to the patient as possible while ensuring that everything in terms of privacy and HIPAA compliance will hold.
It seems like some EHRs don’t necessarily like sharing information present in their systems with other EHR providers. Do you see any obstacles related to this and if so, how will you manage them?
Absolutely, there will always be situations like that. But, one thing that most EHRs are okay with is pushing the data out. However, they don't like incoming data because now you're manipulating their records in their system. That's where we aren't going to act. So, everything I mentioned earlier involves only pulling the data from the EHR systems. For example, you want to use your health insurance company's website to look up providers in your network. If all we do is take that login page and integrate it into the app and authenticate it only once so that it doesn't show up multiple times, all we're doing is connecting you to the insurance page, but you are essentially skipping several steps and also, you're ultimately accessing everything through them. So, by doing this with the EHRs, we don't intend to write anything into their systems or copy anything out of their systems. We are simply trying to give you access to their system without you being conscious of it and going through the hassle of logging in to every provider’s system every time. So, our goal is to avoid any incoming traffic to these EHRs as much as possible, and at this point in time we do not see any reason to do otherwise. But you do have a valid point, especially as SMIRTA grows and EHRs do see some kind of competition which they don't want. At that time, they can even disallow us from reading the information so we will have to work through those hurdles and come up with solutions as they arise.
What would you say the biggest hurdle or bottleneck for SmartER Care is right now?
For SmartER Care, the biggest hurdle is getting the end users and the patient base, as we spoke about earlier. Ultimately, whether we work through health systems or employers/ insurance companies, they will find value for this product if there are enough people using it. So, our biggest challenge will be getting this onto as many devices as we can.
Do you have any advice for medical professionals or students as you see the technology landscape changing medicine?
The key thing we always run into as we are dealing with MDs is that lots of time, technology scares them. So, my first thing is not to be scared of technology because it's never going to replace physicians, so you shouldn't even worry about that. Secondly, it's only going to make their lives easier. Of course, there is a learning curve which is present whether you are a physician or not. In any industry, when you are trying to embrace new technology or anything that's out of your comfort zone, it will take effort. But, more often than not when you embrace that, you realize that it pays off. So, the tradeoff in getting out of your comfort zone and embracing the technology is significantly going to help not only physicians, but also their patients and everyone they work with. On the other hand, I also know many physicians who love technology and are very entrepreneurial and they go out of their way to see how they can improve the process of using technology in order to help their health systems. So, if you are one of them who is scared of technology, don't be. I think my wife is a great example. She has an iPhone on which she only uses 2 things, whereas she could be using about 200 things, so, I see it first-hand.
Is there in field of medicine in which you see technology playing a larger role in the future?
It's difficult to say right now, but I think AI will drive a lot of factors across different aspects of healthcare. I currently cannot think of any specific part of the healthcare industry which needs to be targeted more than any other. However, across the board, I think AI will play a significant role in the next 10-20 years at least. When I say healthcare, I’m not only referring to the purely clinical aspects either. AI can be process oriented. It can literally be anything related to a patient and the hospital.